JP6349096B2 - Tool inspection method and tool inspection apparatus - Google Patents

Description

  The present invention relates to a tool inspection method and a tool inspection apparatus.

  In Patent Document 1, as a method of inspecting a chip of a tool tip, a tool of a milling machine is transferred by a traveling robot and fixed to a jig on a tip inspection / exchange work table, and slit light is emitted from a light projection window of a structure light unit. It is disclosed that the tool tip is projected, the tool tip is photographed with a camera, analyzed by an image processing apparatus, and index data representing the size of the tool tip defect is obtained to determine the necessity for replacement. Regarding the necessity of tool tip replacement, the Y-axis direction and Z-axis direction spread of the recess formed by wear progressing from the edge portion of the tool tip is recognized as index data, and the index data is a preset allowable value. It is determined that there is a need for replacement when the value exceeds.

JP-A-10-96616

  In the method described in Patent Document 1, it is necessary to position the tool tip at the inspection position when photographing the tool tip with a camera when measuring the spread in the Y-axis direction and the Z-axis direction of the recess formed by wear. is there. That is, in order to accurately determine whether the tool tip needs to be replaced, it is necessary to accurately position the tool tip at the inspection position.

  Therefore, in the method described in Patent Document 1, the accuracy of the tool tip replacement necessity is affected by the positioning accuracy of the tool tip. Further, since it takes time to position the tool tip, it takes time to determine whether the tool tip needs to be replaced.

  The present invention has been made in view of the above problems, and an object thereof is to provide a tool inspection method and a tool inspection apparatus capable of inspecting a tool with a simple method with high accuracy.

  The tool inspection method according to the present invention includes a reference image acquisition step of acquiring a reference image by photographing a reference tool as a reference, and an inspection target image acquisition step of acquiring an inspection target image by capturing a target tool to be inspected A correlation value calculating step of calculating a correlation value between the reference image and the inspection target image based on a phase component obtained by frequency-decomposing the reference image and the inspection target image, and based on the correlation value A determination step of determining the degree of deterioration of the target tool.

  A tool inspection apparatus according to the present invention includes an inspection target image acquisition unit that acquires an inspection target image by imaging a target tool to be inspected, and a phase obtained by frequency-resolving a reference image of a reference tool and the inspection target image A correlation value calculation unit that calculates a correlation value between the reference image and the inspection target image based on a component, and a determination unit that determines a degree of deterioration of the target tool based on the correlation value. And

  According to the present invention, the degree of deterioration of the target tool is determined based on the correlation value between the reference image calculated based on the phase component obtained by frequency decomposition of the reference image and the inspection target image and the inspection target image. The inspection can be performed without positioning the target tool to be inspected. Therefore, the tool can be inspected with high accuracy by a simple method.

It is a schematic block diagram of the tool inspection apparatus which concerns on embodiment of this invention. It is a flowchart which shows the procedure of the tool inspection method which concerns on embodiment of this invention. The image (reference | standard image) of the blade edge | tip of an unused throw away tip is shown. It is an image of the cutting edge of the throwaway tip (image to be inspected), and is an unused image. It is an image (inspection target image) of the blade tip of the throw-away tip, and is an image after 100 times of use. It is an image (inspection target image) of the blade tip of the throw-away tip, and is an image after being used 200 times. A reference image is shown. It is an image to be inspected and is the same image as the reference image. This is an image to be inspected and is an image whose position is shifted from the reference image. It is a graph showing the phase-only correlation function when the phase-only correlation method is applied to the same two images in which the displacement amount of the subpixel is (δ ₁ , δ ₂ ) = (d, 0). It is a graph which shows the approximate correlation value of a subpixel level. The reference | standard image and inspection object image of the throwaway chip | tip 1,2,3 in 1st Example are shown. It is a graph which shows the change of the correlation value by the phase only correlation method of the reference | standard image of throwaway chip | tip 1,2,3, and a test object image. It is a graph which shows the change of the correlation value by the normalization cross-correlation method of the reference | standard image of throwaway chip | tip 1,2,3, and a test object image. The reference | standard image and inspection object image of throwaway chip | tip 1,2,3 in 2nd Example are shown. It is a graph which shows the change of the correlation value by the phase only correlation method of the reference | standard image of throwaway chip | tip 1,2,3, and a test object image. It is a graph which shows the change of the correlation value by the normalization cross-correlation method of the reference | standard image of throwaway chip | tip 1,2,3, and a test object image.

<First Embodiment>
First, a first embodiment of the present invention will be described with reference to FIGS.

  The tool inspection apparatus 100 according to the first embodiment of the present invention is an inspection apparatus that measures the degree of tool deterioration. In this embodiment, a case where the inspection target is a throw-away tip of various tools will be described.

  As shown in FIG. 1, the tool inspection apparatus 100 is acquired by a camera 1 as an inspection target image acquisition unit that acquires an inspection target image by photographing a cutting edge of a throw-away tip mounted on a processing apparatus. And a computer 2 that performs image processing of the image data and determines the degree of deterioration of the throw-away chip. The camera 1 is attached to a processing apparatus. The computer 2 may be provided adjacent to the processing device, or may be provided at a location away from the processing device.

  The computer 2 includes a display 21 as a display unit capable of displaying image data, and a keyboard 22 and a mouse 23 as input units capable of inputting an instruction from a user. The computer 2 stores an image of the cutting edge of an unused throw-away tip as a reference tool as a reference image. The computer 2 determines the degree of deterioration of the throw-away chip that is the inspection target based on the comparison between the inspection target image captured by the camera 1 and the reference image stored in the computer 2 in advance. Specifically, the correlation value calculation unit calculates a correlation value between the inspection target image and the reference image, and the determination unit determines the deterioration degree of the throw-away chip based on the correlation value.

  Next, the throw-away tip inspection method will be described in detail.

  As preparation before the inspection, the cutting edge of an unused throw-away tip is imaged in advance, and a reference image serving as a reference for determining the degree of deterioration of the inspection target is acquired (reference image acquisition step). The acquired image is stored in the computer 2.

  The throw-away tip is inspected in the procedure of steps 11 to 14 shown in FIG.

  In step 11, it is determined whether or not the number of uses of the throw-away tip has reached a predetermined number of times set in advance. The number of times the throwaway tip is used can be obtained by counting the number of workpieces processed using the throwaway tip. Instead of this, it may be determined whether or not the usage time of the throw-away tip, that is, the time for processing the workpiece using the throw-away tip has reached a predetermined time.

  If it is determined in step 11 that the number of times the throwaway tip has been used has reached a predetermined number, the process proceeds to step 12.

  In step 12, an image of the tip of the throw-away tip is captured by the camera 1 to acquire an inspection target image (inspection target image acquisition step).

  In step 13, the inspection object image acquired by the camera 1 is read by the computer 2, and the computer 2 calculates a correlation value between the reference image and the inspection object image (correlation value calculation step). The correlation value is a value indicating the similarity between the reference image and the inspection target image. The correlation value is calculated using the phase only correlation method. The phase only correlation method is a correlation method using only phase components having shape information of an image obtained when Fourier transform is performed on the image. That is, this is a correlation method that normalizes the amplitude component that is the luminance information of the image and uses only the phase component that is the shape information of the image. As described above, in step 13, the correlation value between the reference image and the inspection target image is calculated based on the phase component obtained by frequency decomposition of the reference image and the inspection target image.

The two images to be compared are f ₁ (x, y) and f ₂ (x, y), their Fourier transform is F ₁ (u, v), F ₂ (u, v), and the phase component obtained from the Fourier transform Is F ₁ ′ (u, v) and F ₂ ′ (u, v), the phase-only correlation function g (x, y) is obtained by inverse Fourier transform of the following G (u, v). Incidentally, F2 * denotes the complex conjugate of F _2.

  Since the phase only correlation method uses only shape information that is a phase component of an image, the phase only correlation function shows a sharp peak when applied to two similar images. The degree of similarity is evaluated using the maximum value among the plurality of peaks of the phase-only correlation function as the correlation value. When the phase only correlation method is applied to the same image, the correlation value becomes 1.0, and when the phase only correlation method is applied to a similar image, the correlation value changes according to the similarity of the images. That is, when the phase-only correlation method is applied to the reference image and the inspection target image, the correlation value becomes a value close to 1.0 when the degree of deterioration of the throw-away chip is small, and becomes 0 as the deterioration progresses. A close value.

  FIG. 3A shows a reference image that is an image of the blade tip of an unused throw-away tip, and FIGS. 3B, 3C, and 3D show an image to be inspected that is an image of the blade tip of the throw-away tip. 3B, 3C, and 3D are images that are unused, after 100 times of use, and after 200 times of use, respectively.

  Table 1 shows correlation values between the reference image and the inspection target image obtained by applying the phase-only correlation method to the reference image shown in FIG. 3A and the inspection target images shown in FIGS. Table 1 also shows, as a comparative example, a reference image obtained by applying a conventional cross-correlation method to the reference image shown in FIG. 3A and the inspection object images shown in FIGS. The correlation value with the image is also shown. The normalized cross-correlation method is a correlation method that uses both an amplitude component that is luminance information of an image and a phase component that is shape information of the image.

  As can be seen from Table 1, the correlation value obtained by the phase-only correlation method decreases as wear of the throw-away tip progresses after 100 times of use and 200 times of use. In addition, since the correlation value obtained by the phase-only correlation method is lower than the correlation value obtained by the normalized cross-correlation method, the phase-only correlation method is superior in sensitivity for evaluating the degree of similarity compared to the normalized cross-correlation method. It can be said. Thus, since the phase only correlation method is specialized for the phase component which is shape information, the correlation value decreases with high sensitivity according to the shape change accompanying wear of the throw-away tip. Therefore, it can be said that the similarity between the reference image and the inspection object image can be accurately evaluated by using the correlation value obtained by the phase only correlation method.

  FIG. 4A shows a reference image, and FIGS. 4B and 4C show inspection target images. The inspection target image shown in FIG. 4B is the same image as the reference image, and the inspection target image shown in FIG. 4C is an image whose position is shifted from the reference image.

  Table 2 shows correlation values between the reference image and the inspection target image obtained by applying the phase-only correlation method to the reference image shown in FIG. 4A and the inspection target images shown in FIGS. 4B and 4C. Further, Table 2 shows, as a comparative example, a reference image obtained by applying the normalized cross-correlation method, which is a conventional method, to the reference image shown in FIG. 4A and the inspection object images shown in FIGS. 4B and 4C and the inspection object. The correlation value with the image is also shown.

  As can be seen from Table 2, the correlation value obtained by the normalized cross-correlation method is lower than 1.0 when the inspection target image is a misalignment image, and the reference image and the inspection target image cannot be recognized as the same image. On the other hand, the correlation value by the phase only correlation method is 1.0 even if the inspection target image is a misaligned image, and the misaligned inspection target image is recognized as the same image as the reference image. This is because, in the phase-only correlation method, an image misalignment can be considered as a phase component misalignment. Therefore, if there is an image misalignment, a correlation peak misalignment is generated by the misalignment of the image. This is because the reference image and the reference image can be recognized as the same image. At this time, since the value of the correlation peak does not change, it is not necessary to correct the positional deviation. Therefore, if the phase only correlation method is used, the similarity between the reference image and the inspection object image can be evaluated without aligning the reference image and the inspection object image.

  As described above, by using the correlation value obtained by the phase-only correlation method, it is not necessary to align the reference image and the inspection target image, and the similarity between the two can be accurately evaluated.

  Returning to FIG. 2, in step 14, the degree of deterioration of the throw-away chip based on the correlation value calculated in step 13, that is, the correlation value obtained by applying the phase only correlation method to the reference image and the inspection target image. Is determined (determination step). For example, when the correlation value calculated in step 13 decreases to a predetermined value, the throw-away tip is determined to have a lifetime. Further, deterioration levels 1 to 5 are set in advance according to the magnitude of the correlation value, and when the deterioration level 5 is reached, the throw-away tip is determined to have a lifetime.

  The above-described image processing of the reference image and the inspection target image and the determination of the degree of deterioration of the throw-away chip are automatically executed by software stored in the computer 2. As a result, if it is determined that the throw-away tip has reached the end of its life, a notification that prompts replacement is issued.

  According to the above 1st Embodiment, there exists an effect shown below.

  Since the degree of deterioration of the target tool is determined based on the correlation value between the reference image calculated based on the phase component obtained by frequency decomposition of the reference image and the inspection target image and the inspection target image, the target to be inspected There is no need to position the tool, and there is no need to align the reference image and the inspection object image. Therefore, the tool can be inspected with high accuracy by a simple method.

  In the past, skilled techniques were required to determine tool life. When not relying on skilled technology, tools were changed at a safe number of uses. However, according to the present embodiment, it is possible to automatically determine the tool life simply by photographing the cutting edge of the tool, so that no skill is required, and the tool can be used up to the true life. And cost can be reduced.

Second Embodiment
Next, a second embodiment of the present invention will be described with reference to FIGS. Hereinafter, only differences from the first embodiment will be described.

  The similarity evaluation based on the correlation value calculated using the phase only correlation method can be evaluated with high accuracy when the reference image and the inspection target image are shifted by integer pixels. However, when there is a sub-pixel level shift between the reference image and the inspection target image, that is, a shift of less than one pixel, the accuracy of similarity evaluation is lowered. For example, when the phase-only correlation method is applied to the same two images that are shifted by 25 pixels in one direction, the correlation value becomes 1.0, whereas the same two images that are shifted by 25.5 pixels in one direction. When the phase only correlation method is applied to an image, the correlation value is reduced to about 0.6. Thus, the similarity evaluation based on the correlation value calculated using the phase only correlation method is an evaluation at the pixel level.

  In the second embodiment, a correlation value calculation method between the reference image and the inspection target image is used so that the similarity between the two images can be accurately evaluated even when the reference image and the inspection target image have a sub-pixel level shift. Is different from the first embodiment. This will be described in detail below.

The phase-only correlation function g (n ₁ , n ₂ ) obtained by the phase-only correlation method can be approximated by the following equation using the sinc function. δ ₁ and δ ₂ are sub-pixel level shifts in the X-axis and Y-axis directions (see FIG. 3A) of the image, respectively, −0.5 ≦ δ ₁ ≦ 0.5, −0.5 ≦ δ ₂ ≦ 0 .5.

FIG. 5 is a graph showing a phase-only correlation function when the phase-only correlation method is applied to the same two images in which the displacement amount of the subpixel is (δ ₁ , δ ₂ ) = (d, 0). . In FIG. 5, the solid line is discrete data that appears as a peak. The correlation value between the two images is about 0.85, which is the maximum value among the plurality of peaks, and is smaller than the actual correlation value 1.0.

  Assuming that the coordinate corresponding to the actual correlation value 1.0 is 0 and the amount of deviation between the coordinate and the peak corresponding to the peak indicating the maximum correlation value 0.85 is d (pixels), the peak value near the maximum correlation value is The coordinates are 1 + d, 1-d, 2 + d, 2-d as shown in FIG.

  The correlation value when there is a subpixel level shift between the reference image and the inspection target image is approximated to the total value of the correlation values of coordinates d, 1 + d, 1-d, 2 + d, and 2-d. That is, the correlation value at the sub-pixel level is approximated to an addition value of the maximum correlation value of the phase-only correlation function and the correlation value in the vicinity of the maximum correlation value. This will be described.

  For the coordinates d, 1 + d, 1-d, 2 + d, 2-d shown in FIG. 5, the correlation value is calculated using the sinc function of the following equation.

  FIG. 6 shows a graph in which all of the five correlation values are added together in -0.5 ≦ d ≦ 0.5, which is the entire range of d. As can be seen from FIG. 6, since the approximate correlation value is about 1.0 in the entire range of −0.5 ≦ d ≦ 0.5, the correlation value at the subpixel level is the maximum of the phase-only correlation function. It can be seen that approximation is possible with the sum of the correlation value and the correlation value near the maximum correlation value.

  The above calculation of the sub-pixel level correlation value is performed in step 13 in the flowchart of FIG.

  In the present embodiment, the case where the correlation value at the sub-pixel level is calculated by adding a total of five correlation values of the maximum correlation value and the four correlation values in the vicinity centering on the maximum correlation value has been described. Alternatively, the sub-pixel level correlation value may be calculated by adding a total of three correlation values, that is, the maximum correlation value and two neighboring correlation values centered on the maximum correlation value. That is, three correlation values of coordinates d, 1 + d, 1-d may be added. Further, seven correlation values of coordinates d, 1 + d, 1-d, 2 + d, 2-d, 3 + d, and 3-d may be added. As described above, the correlation value at the sub-pixel level is calculated by adding the maximum correlation value of the phase-only correlation function and the correlation value near the maximum correlation value.

  According to the above 2nd Embodiment, there exists an effect shown below.

  Even when there is a sub-pixel level shift between the reference image and the inspection target image, by calculating the correlation value by adding the maximum correlation value of the phase-only correlation function and the correlation value in the vicinity of the maximum correlation value, It is possible to accurately evaluate the similarity of the inspection target image.

<First embodiment>
The first embodiment will be described with reference to FIGS.

  Using the same type of throwaway tips 1, 2, and 3 in different usage environments, the reference image of the blade tip of the unused throwaway tips 1, 2, and 3, and the throwaway tips 1, 2, and 3 according to the number of uses The degree of deterioration of each throw-away tip 1, 2, 3 was evaluated by calculating the correlation value with the inspection target image of the blade edge.

  The correlation value was calculated by the phase only correlation method and also by the normalized cross correlation method as a comparative example. The correlation value by the phase only correlation method was calculated by the method shown in the second embodiment.

  FIG. 7 shows reference images and inspection target images of the throw-away chips 1, 2, and 3. The inspection target image is an image in an unused state from the left in the drawing, after being used 100 times, after being used 200 times, and after being used 250 times. FIG. 8A is a graph showing a change according to the number of times of use of the correlation value by the phase-only correlation method between the reference images of the throw-away chips 1, 2, and 3 and the inspection target image, and FIG. , 2 and 3 are graphs showing changes according to the number of times of use of the correlation value by the normalized cross-correlation method between the reference image and the inspection target image. 8A and 8B, the correlation values of the throw-away chips 1, 2, and 3 are indicated by a solid line, a broken line, and a one-dot chain line, respectively. The calculation of the correlation value was performed for each use.

  As can be seen from FIG. 8A, the correlation value by the phase-only correlation method converges to 0 as the number of uses increases. Further, since the correlation value at the sub-pixel level is calculated, the noise is smoothly converged to 0. On the other hand, as can be seen from FIG. 8B, the correlation value obtained by the conventional normalized cross-correlation method does not converge to one value due to a positional shift. Therefore, it can be said that the phase only correlation method is effective for evaluating the degree of deterioration of the throw-away chip.

<Second embodiment>
A second embodiment will be described with reference to FIGS.

  Using the same type of throwaway tips 1, 2, and 3 in different usage environments, the reference image of the blade tip of the unused throwaway tips 1, 2, and 3, and the throwaway tips 1, 2, and 3 according to the number of uses The degree of deterioration of each throw-away tip 1, 2, 3 was evaluated by calculating the correlation value with the inspection target image of the blade edge. Further, the wear rate of the throw-away tips 1, 2, and 3 was changed. Specifically, with reference to the wear rate of the throw-away tip 1, adjustment was made so that the wear rate of the throw-away tip 2 was fast and the wear rate of the throw-away tip 3 was slow.

  The correlation value was calculated by the phase only correlation method and also by the normalized cross correlation method as a comparative example. The correlation value by the phase only correlation method was calculated by the method shown in the second embodiment.

  FIG. 9 shows reference images and inspection target images of the throw-away chips 1, 2, and 3. The inspection target image is an image in an unused state from the left in the drawing, after being used 100 times, after being used 200 times, and after being used 250 times. FIG. 10A is a graph showing a change according to the number of times of use of the correlation value by the phase-only correlation method between the reference images of the throw-away chips 1, 2, and 3 and the inspection target image, and FIG. 10B shows the throw-away chip 1 , 2 and 3 are graphs showing changes according to the number of times of use of the correlation value by the normalized cross-correlation method between the reference image and the inspection target image. 10A and 10B, the correlation values of the throw-away chips 1, 2, and 3 are indicated by a solid line, a broken line, and an alternate long and short dash line, respectively. The calculation of the correlation value was performed for each use.

  As can be seen from FIG. 10A, the rate of decrease of the correlation value varies depending on the wear rate. From this, it can be said that if the correlation value by the phase-only correlation method is used, it is possible to accurately evaluate the life due to the difference in wear rate.

  The embodiment of the present invention has been described above. However, the above embodiment only shows a part of application examples of the present invention, and the technical scope of the present invention is limited to the specific configuration of the above embodiment. Absent.

1 Camera (image acquisition unit for inspection)
2 Computer 100 Tool inspection device

Claims (4)

A reference image acquisition step of capturing a reference image by photographing a reference tool as a reference;
An inspection object image acquisition step of acquiring an inspection object image by imaging a target tool to be inspected;
A correlation value calculating step of calculating a correlation value between the reference image and the inspection target image based on a phase component obtained by frequency-decomposing the reference image and the inspection target image;
A determination step of determining the degree of deterioration of the target tool based on the correlation value;
A tool inspection method comprising:   The tool inspection method according to claim 1, wherein in the correlation value calculation step, the correlation value is calculated by applying a phase-only correlation method to the reference image and the inspection target image.   The correlation values calculated in the correlation value calculation step are the maximum correlation value of the phase only correlation function obtained by applying the phase only correlation method to the reference image and the inspection target image, and the correlation in the vicinity of the maximum correlation value. The tool inspection method according to claim 2, wherein the calculation is performed by adding the value. An inspection target image acquisition unit that acquires an inspection target image by imaging a target tool to be inspected;
A correlation value calculation unit for calculating a correlation value between the reference image and the inspection target image based on a phase component obtained by frequency-resolving the reference image of the reference tool and the inspection target image;
A determination unit that determines the degree of deterioration of the target tool based on the correlation value;
A tool inspection apparatus comprising: